Towards Mitigating Privacy Concerns on Fall Detection Techniques for Elderly / Physically challenged

Thursday 2:40pm - 3:00pm, (CITRENZ 1 Room)

Elderly and some physically challenged people require continuous attention and care due to their physical and health conditions. Significant proportion of these people are prone to fall-based critical injuries ranging from minor to fatal (Zhang et al., 2015). The definition of a fall is an event that results in a person coming to rest unintentionally and abruptly on the ground or other low surfaces (Pannurat et al., 2017). The World Health Organization (WHO) stated that about 646,000 fatal falls happen each year across the world and most of the sufferers are more than 65 years old (World Health Organization, 2018). Moreover, globally there is a significant shortage of caregivers to take care of the growing ageing population and it’s no different in New Zealand. Technology has come to the aid and multiple sensory and artificial intelligence (AI)-based techniques for automatic fall detection and warning system have been proposed by the research fraternity as part of the ambient-assisted living.

The different technology-based fall detection techniques can be categorized into various categories (Fig.1). Most of these either need sensors or equipment that can work with radio frequency or can be connected to through WiFi or cellular connections. For instance, context-based fall detection uses equipment like surveillance cameras and efficient and low-cost computer vision techniques to detect falls. These indoor security cameras do not affect the day to day privacy of users.

A primary concern for users, however, are most of these technology-based fall detection systems now are cloud-based and have online connectivity, which make them vulnerable to data privacy issue. Lot of these fall monitoring and detection systems rely on AI techniques.

Sayan Kumar Ray